Decreasing Diagnoses of Birth Asphyxia in Canada: Fact or Artifact

We assessed temporal trends in birth asphyxia in Canada, to determine whether changes were real or secondary to changes in coding.We used data from the Canadian Institute for Health Information Discharge Abstract Database to study the national incidence of birth asphyxia, by using International Classification of Diseases codes. We also studied birth asphyxia by using data from the Nova Scotia Atlee Perinatal Database. In the Nova Scotia Atlee Perinatal Database, we defined a case of birth asphyxia as a live birth with an Apgar score at 5 minutes of < or =3, depression at birth requiring resuscitation with a mask for > or =3 minutes and/or intubation, or neonatal postasphyctic seizures.Nationally, between 1991 and 2005, the incidence of birth asphyxia decreased significantly, from 43.8 to 2.4 cases per 1000 live births. The rate of decrease was highest between 1991 and 1998, corresponding to a period when strict Canadian and international criteria for the diagnosis of birth asphyxia were published. By comparison, neither national rates of related diagnoses nor Nova Scotia birth asphyxia rates, which ranged from 8.8 to 14.3 cases per 1000 live births, showed evidence of a decrease during the study period.Comparisons of national trends in birth asphyxia diagnoses and trends in conditions associated with birth asphyxia, both nationally and in Nova Scotia, suggest that the dramatic decrease in the diagnosis of birth asphyxia is an artifact of changes in the use of International Classification of Diseases coding associated with the publication of stricter diagnostic definitions of birth asphyxia. We conclude that International Classification of Diseases codes are not useful for surveillance of birth asphyxia.